As the Mets return home to Shea for game six, they face the threat of elimination at the hands of Cardinal ace Chris Carpenter. The Mets threadbare rotation turns to rookie John Maine, who last just four innings in a 9-6 game two loss.

It may be an understatement to say that it doesn't look good for the Mets. They have two win the last two games of the series at home, and they don't even know who the starter might be tomorrow if they end up winning tonight.

But given the general oddness of this entire postseason, I am reluctant to count the Mets out just yet.

I would rather look at the historical record to determine what it says about the Mets' odds. I looked up all seven-game playoffs series in which the team trailing three games to two played hosts for game six.

There were 82 such series. Of these, the trailing team won game six 46 times. That translates into a .561 winning percentage. So the odds are in the Mets favor for a win tonight, but what about the entire series?

Of the 46 teams that won game six, 26 won game seven. (Note that one series switched hosts for game six and seven, the 1912 World Series, which the once 3-2 leading Red Sox won. I choice to keep them in the study even though game seven was in Boston.)That's an even better winning percentage, .565.

So it looks like the odds are in the Mets favor, right? Well, no.

They have a slightly better than a 50-50 shot in each game, but overall they have about a 1-in -3 chance (31.71% to be exact). That's considerably better than the 1-in-4 chance they would have if all things were equal (i.e., 50-50 for each game).

My next question is if the winning percentages in game six and seven above are just a result of having a home field advantage. What is the home field advantage in the postseason anyway?

I looked it up. From the 1903 World Series through this year's ALCS (but omitting the yet complete NLCS), the home team has won 625 games while the visitors have won 534. That's a .539 winning percentage, which is almost midway between the percentages we see in game six and seven when the home team initially trails two games to three.

It seems that for some reason the added pressure almost doubles the home field affect. It may have to do with the fact that the home teams in game six and seven are typically the higher seeded team, at least in intraleague series. However, that has no bearing on World Series games which until very recently alternated the homefield for games six and seven between the two leagues.

So there may be something to a slight advantage for the Mets in each of these two games. But remember that the overall effect is still in the Cardinals favor.

I hope that clears it up.

By the way, here are the 26 series in which the team trailing two games to three went home in game six and won the series. The last instances were both LCS's in 2004: